The last thing that users of broken-down air-conditioning or heating units want to do is wait.
Today's heating, ventilation and air-conditioning (HVAC) business is centered on fixing customers' problems immediately. Lennox Residential, one of the biggest players in that sector, configures its distribution network so that large numbers of products and parts are positioned for immediate pickup or same-day delivery at locations around the country. Even parts that aren't stocked locally can usually be delivered by the next morning.
Complicating matters is the industry’s high degree of unpredictability, not to mention the built-in seasonality of a business that provides relief from winter’s cold and summer’s heat.
Lennox Residential is a unit of Lennox International, a leading provider of climate-control systems for both home and business. The former accounts for roughly 43 percent of the parent company’s $5bn in U.S. sales.
For a number of years, Lennox Residential had served its nationwide customer base through a pair of distribution centers near Des Moines, Iowa – one for products, the other for parts. The facilities supported 66 Lennox-owned retail stores.
Some 98 percent of SKUs account for 62 percent of Lennox Residential’s revenues. Many are slow-movers that are subject to highly intermittent demand. At the same time, Lennox must cope with both the new and the old: nearly half of a line of finished goods might be replaced by new models in a given year, and the company pledges to service existing units for 15 years or more.
Lennox was aiming to boost the number of SKU locations for goods and spare-parts from 450,000 to more than 700,000. All of this, of course, was to be accompanied by optimized inventories, balanced allocations and a marked improvement in service levels. The driving goal was “to grow revenue at a cost we could afford,” says Keith Nash, vice president of supply chain logistics.
The solution lay in the transition to a hub-and-spoke model of distribution, accompanied by a significant bump in the number of sites. Lennox ended up creating eight regional and 19 local D.C.s, serving more than 180 stores and supporting five manufacturing facilities. (A ninth regional D.C. was set to open in November of this year, and the company was planning on adding 30 retail stores in 2016). The new network allows Lennox to reach 95 percent of demand by next morning, 98 percent by next day and 50 percent by same-day delivery when necessary.
Faster Response Times
The manufacturing end had to adjust operations accordingly. Previously, it took six days from the time that demand was identified to the moment when a factory would change its schedule. Now that happens overnight, says Nash.
The new network would be unsupportable without a corresponding upgrade of forecasting and fulfillment techniques. Lennox needed to embrace advanced analytics that could model and forecast demand in a highly unpredictable environment.
Setting aside traditional forecasting methods, which rely heavily on historical demand, Lennox embraced demand modeling. With the help of the SO99+ application from ToolsGroup Inc., it adopted a “bottom-up” approach, which draws on unique daily demand patterns for each individual SKU-location combination. (A tall order, considering that Lennox was dealing with more than 110,000 such pairs throughout the supply chain.)
Instead of getting locked into a single number for future demand, the process incorporates those daily patterns into a more reliable statistical forecast, accompanied by a confidence interval. The method allows Lennox to drill down into the most granular level of detail possible when setting target stocks.
The addition of external and downstream channel data adds an element of “outside-in” forecasting to the demand model. In the process, downstream demand gets translated into a demand signal for each upstream SKU-location combination, reducing the demand-latency gap from weekly to daily.
“By adding downstream channel data, we began to use market-driven demand to improve market responsiveness … and make more optimal trade-offs to maximize key business outcomes such as customer service, profitability and revenue,” according to a Lennox paper on the transformation.
The final phase of the project utilized machine learning to address patterns of seasonality, “enabling us to continuously fine-tune the signal-to-noise ratio,” Lennox said. As a result, it was able to identify patterns and trends that weren’t visible through standard statistical approaches.
A Complex Calculation
With the help of SO99+, Lennox is now able to model demand patterns related both to seasonality and the variable consumption that was so difficult to forecast previously. It can determine the optimal mix of inventory and service levels down to the store level, accounting for such factors as delivery frequency, order size and distance to be traveled. Demand simulations allow it to see how various scenarios might play out prior to implementation. And it’s able to pre-build inventories in anticipation of seasonal peaks.
The ToolsGroup application has yielded a number of direct benefits. Compared with 2008, Lennox has half the number of people working on planning and supplier management, with double the number of purchase-order lines. And forecast accuracy is approaching 85 percent, compared with the mid-50s previously.
On the inventory side, Lennox has reduced overall volumes by nearly 20 percent, despite a 250-percent increase in physical locations. Inventory turns rose by 39 percent between 2011 and 2015. The facing fill rate is up 25 percent, with an increase in the number of lines filled. And distribution costs as a percentage of sales have dropped by more than 15 percent.
Even more important is the resulting boost in customer-service levels. Lennox has improved its ability to provide same-day delivery by 40 percent, while upping the number of orders capable of being delivered the next morning from 35 percent to a commanding 98 percent. Revenue growth from the time the system went live was approaching 50 percent, and Lennox Residential’s market share rose by nearly 25 percent over that period. “A lot of that was attributable publicly to our supply chain,” says Nash.
Enjoy curated articles directly to your inbox.